DAEMONitor: Monitoring System

DAEMONitor started in 2022 as a way to watch pm2-managed processes across separate cloud and bare-metal instances. It has since grown into a general infrastructure monitoring platform: a monitoring agent runs on each system and loads plugins that collect metrics, and a Nuxt dashboard renders the whole estate.

The agent is built around a BasePlugin → MonitoringPlugin inheritance chain with a setup → monitor → refresh → teardown lifecycle, so a new class of thing to watch is a plugin rather than a fork. Current plugins cover OS metrics, pm2, Docker containers, LXC, MongoDB, website/origin checks, WordPress posture, network scanning, log watching, SNMP devices, TP-Link extenders, 3CX PBXs and branch SBCs, eWeLink, and Cloudflare. Cloudflare inverts the model — a scheduled Worker pushes Pages projects, Workers, and zone data into the API using a reporting key, which keeps the account API token in the Worker environment instead of in every client config.

The dashboard groups related systems into stacks and groups with dedicated detail screens that render every member's metrics inline, splits network throughput into separate IN and OUT series everywhere it appears, surfaces per-interface link state on SNMP devices, and explains why an account is degraded rather than just flagging it. Tenancy is handled through organization membership as an access path, with org-scoped API keys for customer-facing portals. Built on Nuxt 4, Vue 3, Nuxt UI and Pinia with Supabase, plus native iOS and Android builds via Capacitor with Firebase push.